Analysis of Traffic Operations
for
the Movement of Very Large Vehicles
on the
Edmonton-Fort McMurray High- Wide Load Corridor
by
John Morrall
and
Walid Abdelwahab
Department of Civil Engineering
The University of Calgary
and
Al Werner
Alberta Transportation and Utilities
Acknowledgements
This study was made possible through the support of Alberta Transportation and
Utilities.
Analysis of Traffic Operations for the Movement of Very Large Vehicles on
the Edmonton-Fort McMurray High-Wide Load Corridor
Abstract
Alberta Transportation and Utilities, Alberta Power,Ltd. and Trans Alta Utilities Ltd.
in conjunction with Syncrude Canada Ltd. and other contractors have developed a highwide corridor from Edmonton to Mildred Lake (north of Fort McMurray). The corridor
allows the transport of very large vehicles up to 8 m wide, 8 m high and 27 m long. The
transport weights allowed on the high-wide corridor ranges from 122 tonnes in the summer
months to 177 tonnes in the winter months. The very large vehicles allowed on the highwide corridor transport process modules, dressed equipment and major pre-assemblies for
the oil sands extraction plants in the Fort McMurray area.
The purpose of this paper is to discuss the results of a recently completed analysis of
the impact of very large vehicles on rural highway traffic operations.
The problem of determining the impact of very large, slow-moving vehicles on twolane highway level of service cannot be solved using conventional highway capacity analysis
procedures. Thus traffic simulation models were utilized to determine the impact of 1600
very large slow-moving vehicles and up to 20 % heavy trucks on the traffic stream.
Also discussed in the paper is how the findings of the simulation studies were used to
demonstrate the importance of strategically located passing lanes and turn-outs in helping
to maintain an acceptable level of service on the high-wide corridor during periods of heavy
truck movements.
1
Intro'duction
Alberta Transportation and Utilities, Alberta Power Ltd. and Trans Alta Utilities Ltd.
in conjunction with Syncrude Canada Ltd. and other contractors have developed what is
known as High-Wide Load (HWL) transportation corridors from Edmonton to Mildred
Lake. These corridors, depicted in Figure 1, allow the transport of modules and preassemblies of various sizes up to the following maximum dimensions: width 9.1 m (30 ft.);
height 9.1 m (30 ft.); and length 60.6 m (200 ft.). Typical module weights on these routes
range from 122 tonnes in the summer months to 177 tonnes in the winter months. However,
weights as high as 380 tonnes may be conveyed provided size and spacing requirements of
undercarriage wheel assemblies are met for critical bridge structures.
2
Problem Definition
Although the HWL corridors have been selected to permit the shipment of large modules,
two traffic problems are still created by the over dimension modules and the large volume of
freight and equipment required for any plant expansion. The traffic operational problems
are:
1. the slow moving vehicle problem (module transport); and;
2. the high volume of heavy trucks in the traffic stream.
In addition to the above, there is a range of variable road conditions (road widths and
traffic) prevalent along various segments of the HWL corridors. Therefore, each problem
requires a separate non-conventional technical analysis and solution. These are discussed
separately.
2.1
The Slow Moving Vehicle Problem
The slow moving vehicle (SMV) problem involves the shipment of various size modules
including convoys of up to three modules at speeds up to 50 km/h. Figure 2 shows the
maximum dimensions of a module and Figure 3 the configuration and space requirements
of a typical three module convoy. Passing of modules is controlled by pilot trucks at
locations where adequate pavement width exists. The HWL corridor has road sections
of width ranging from 8 to 13 m (26.5 ft. to 42.5 ft.) and, depending on module width,
passing mayor may not be permitted. Table 1 shows the cases where passing may be
accommodated. Where passing along the road cannot be accommodated, then truck
turnouts have to be utilized-if available. Thus traffic delays are a function of SM"
speed, traffic flow, the number of modules transported, and the number of opportunities
for overtaking which in turn is a function of module and road widths.
The number of modules and pre-assemblies for the proposed OSLO plant expansion
is estimated at approximately 1600 of varying sizes. During the peak period there could
be in the range of 5 to 7 lnodules arriving daily at the plant site. In order to estimate
the impact of module shipment on free moving traffic, a simplified slow moving vehicle
model was developed to estimate delay. The engineering solution to reducing delay is to
widen the road, or alternatively, develop a system of strategically located turnouts to allow
modules to stop temporarily and permit both opposing and following vehicles to clear the
module. Figure 4 shows typical truck turnouts that have been designed for incorporation
into Highway 63.
2.2
The Heavy Truck Problem
The OSLO expansion will also require substantial tonnages of construction equipment and
material most of which will be shipped by truck. It is estimated that during construction
the total weight exclusive of modules, pre-assemblies and local materials of bulk and other
materials is estimated at 155,400 to 171,600 tonnes. During peak shipping periods this
could result in 18 to 22 truck loads per day of bulk and other m.aterials. The heavy
truck problem is essentially one of impacting the level of service in terms of increasing
the percent time spent following by free flowing higher speed traffic and increasing the
demand for overtaking while at the same time reducing the opportunity for overtaking
caused by oncoming traffic. In order to determine the impact of a substantial increase
in the number of heavy trucks on Highway 63, a two-lane highway simulation model was
utilized to compliment conventional analysis procedures.
Alberta Slow Moving Vehicle (SMV) Model
3
The engineering solution for a narrow alignment (worst condition) to the slow moving
vehicle problem is to widen the road or develop a system of strategically placed turnouts or
passing bays. The cost of a typical single turnout is $50,000. Delay also has an associated
cost and, depending on traffic volumes, there is an optimum number of turnouts desired to
reduce delay costs. The SMV model was developed to determine the impact of alternative
turnout spacing on traffic delay.
3.1
Derivation of the SMV Model
Models relating effects of slow moving vehicles to operational aspects of two-lane highways utilize a variety of technical parameters. Field measurements are necessary for such
parameters to be estimated for a particular application. Since most of these parameters
are site-specific, models based on these parameters are not usually transferable except
within the same technological and operational settings. Examples of parameters which
are required by such models are those included in the fundamental flow-speed diagram
(e.g. capacity of the highway section, jam density).
A simple, yet reliable, model was desired for the analysis of the effect of slow moving
vehicles on free flowing traffic on a two-lane highway. The model was not to be based
on any data that required field measurements or experiments. The development of this
simplified model was made possible by the fact that no overtakings in the direction of the
SMV is permitted. The model is based on the following assumptions:
1. Traffic arriving at the study section is Poisson distributed with the mean number of
vehicles expected to arrive in time t is m = Q . t, where Q is the flow rate (veh/h).
This assumption involves all the implicit assumptions for the applicability of the
Poisson distribution such as the randomness of arrivals. This assumption is particularly reasonable to this application since the Poisson distribution is inappropriate
over 600 veh/h (Lay, 1985, p. 346).
2. Slow moving vehicles travel at a constant speed, Vs, and free traffic travel at a
constant speed, V" such that Vs ::; V,.
3. Traffic flow rate is sustainable over the entire journey time.
Using the above assumptions, the model was derived as follows:
The number of vehicles crossing an arbitrary point on the highway1 in a given time
interval, t, may be obtained, at any probability level, using the Poisson distribution:
x
Pt (x)
L e-
x
m
•
mZ
L eQ -t . (Q . t):I:
= _:1:=_1_ _ _ _ _ _:1:=_1_ _ _ _ __
x!
x!
(1)
where,
probability that x vehicles will arrive at any point in a given time
interval, t, at the specified arrival rate, and
flow rate, veh/h.
Q
The above equation contains four variables: Pt(x), Q, t, and x. Given any three
variables, equation (1) can be solved for the unknown variable. Pt(x) was set at 0.95
in the calculations of the expected number of vehicles. Since the Poisson distribution is
discrete, i. e. , x = 0,1,2, ... , it is not possible to solve equation (1) analytically for x or
t. Instead, a numerical method was used whereby a computer program was written to
solve for the desired variable given the values of the other three. For any section of the
highway, it is then possible to calculate, at a given probability level, the arrival time of
all vehicles expected to enter that section while the SMV is still travelling along the same
section. At the same time, any vehicle entering the section at time t will not be affected
by the SMV until it catches up with the SMV after time, t c , where:
(2)
where,
n -
L
=
the number of vehicle under consideration (the first vehicle entering
the section is assigned number 1, the next is number 2, and so on),
and
average length of vehicles, m.
Solving equation (2) for tc:
(3)
Note that the second term of the right hand side of equation (3) can be neglected
without loss of accuracy; it only accounts for the difference in the catch-up times when
no vehicles are following the SMV versus that when there is n vehicles following the SMV
at the same speed, Vs.
The delay experienced by any vehicle travelling along any section is simply the difference between two travel times. The first one, t w , is the travel time spent by a free moving
IThis point may be set, for convenience, at any truck turnout from which the SMV starts its journey_
vehicle following a SMV at a speed Vs. The second travel time, two, is the time spent by
the same free moving vehicle without following a SMV. That is,
Delay
= Dt(x) = tw -
two
(4)
where,
delay experienced by vehicle x, arriving at time t,
travel time w£th SMV being in the section,
travel time w£thout SMV being in the section.
For a section of length, d, and using assumption (2) above, the values tw and two can
be written as:
tw =
d-t·Vs
Vs
d-t·Vs
two = - - - Vf
(5)
Substituting equation (5) into (4) and simplifying, yields:
(6)
where C
= speed reduction ratio =
Vf;,V/l.
Inspection of equation (6) shows that the delay experienced by vehicle x arriving at
time t is equal to zero if and only if it arrives at time t = djVs which is equal to the time
required by the SMV to travel a section of length d. However, this is not the case, since
vehicles can arrive at an earlier time than the time required by the SMV to travel the
entire length of d, and still do not experience any delay. This is because of the catchup time required by a free moving vehicle to start experiencing delay. Then, modifying
equation (6) to include the catch-up time, yields:
d
= C(- -
(7)
- C d
(n - l)L
D t (x ) --t+-'----Vs
Vf
(8)
Dt(x)
t - tc)
Vs
Substituting the value of tc from equation (3) into equation (7), yields:
Finally, the total delay resulting from the SMV trip along a section of length d is:
TD
= L Dt(x) = L
x,t
where,
Xt
t
Xt
c .d
(n - l)L)
( T s - t + Vf
is the number of vehicles arriving in time interval t.
(9)
The value of t can be adjusted in the computer program until the step size (the value
of Xt) is at the desired level. However, the smaller the step size, the more computer time
will be required to solve equation (1), and the more accurate the results will bee The most
accurate estimate of the total delay, however, can be calculated by selecting the minimum
increment of Xt (i. e. Xt = 1). The calculations then proceed as follows:
• First, the arrival time, (t, in equation (9)) is calculated for every vehicle, x = 1, 2,
3, ... N a, where Na is the number of affected vehicles in a particular section.
• Second, Xt is set equal to a vector of ones of a length equal to N a , and n is used as
a counter: n = 1, 2, 3, ... N a •
• Finally, the total delay (TD) can be calculated as:
Na
TD =
Lt;
(10)
i=l
where,
t*1
C .d
(ni - l)L
Vs
VI
-- +
-
ti
index of the ith vehicle (equal 1 for the first vehicle, 2 for the
second, and so), and,
arrival time of the ith vehicle.
3.2
Application of the SMV Model
The SMV model developed in this paper (the ALBERTA Simplified model) was applied
to analyze the operational effects of slow moving vehicles on two-lane highway traffic.
The scope of the model was considered satisfactory for the application in hand, since
geometrics of the highway and the dimensions of the SMV did not allow for passing (i.
e. testing of narrow highway sections) of the SMV except on the specified truck turnouts
along the highway. These turnouts were located along the highway at spacings of 10-30
km. The spacing of two succesive turnouts represents the length of the highway section
over which no passing is allowed (i. e. length of a study section).
The ALBERTA Simplifed model utilizes a variety of input data, as described in the
previous section, and generates two major outputs: total delay (in hours), and the number
of affected vehicles. Before presenting the empirical results, the performance of the model
in predicting these two measures is compared with that of another "similar" model developed by DELFT University in the Netherlands (Botma, 1988). The two models share
the same objective of "analyzing the effects on traffic operation of a slow moving vehicle
on two-lane rural roads". While the ALBERTA model has been developed for the specific
purpose of "estimating the delay experienced by motorists following a SMV on a two-lane
highway with no passing opportunity", the DELFT model contains more sophisticated
treatment of the effects of a SMV on two-lane highway traffic with and without passing
opportunity. Therefore, the DELFT model was adjusted for its "No-Overtaking" module
so that results from the two models could be compared.
Table 2 shows total delay in hours for the set of truck turnout spacings used in this
study as calculated by the two models for two values of flow, 100 and 200 veh/h. Another
output from the simplified ALBERTA model, number of affected vehicles , is shown in
Table 3. Also shown in Table 3 is the average delay per vehicle as estimated by the
two models for a flow rate of 200 veh/h. This value was calculated by dividing the total
delay (in hours) for each section, from Table 2, by the number of affected vehicles, and
multiplying the result by 60' min/h. Figures 5, 6 and 7 are graphical representations of
the comparisons held in Tables 2 and 3 for selected flow rates.
The following notes may be made on the comparison of the two models:
1. The two models differ in the values implemented for some of the operational parameters. These parameters include speed at capacity, jam density, and the capacity of
the highway section.
2. The range of distance over which the output from the two models are compared (1325 km) is rather arbitrary. The DELFT model was developed for use with much
shorter distances (up to 5 km) which are more common in the case of farm tractor
trips along rural roads in the Netherlands (Botma, 1988).
In spite of the above, and other minor differences in the assumptions and values of
parameters utilized in both models, the results show a good "fit" between the predictions
of the two models.
4
Simulation of Heavy Trucks on Highway 63
The heavy truck problem is also essentially one of delays incurred by faster following traffic
and is exacerbated where trucks are heavily laden, and are climbing moderate to steep
grades. In addition, increasing opposing traffic reduces the number of gaps for overtaking
to occur. Poor road geometry (no passing zones) also aggravates the problem.
Conventional highway capacity analysis (1985 HCM) does not incorporate procedures
for analysing road sections with a third lane. Therefore, a simulation model was used to
help solve this particular deficiency.
4.1
Traffic Volume and Composition
The base year (1985) average annual daily traffic (AADT) on Highway 63 is approximately
1000 between the junction with Highway 55 and Fort McMurray. The AADT for the design
year of 1993 is estimated at 1350 based on a 3.8 % growth rate.
The base year traffic composition shown in Table 4 indicates a high percentage of
heavy trucks (approximately 20 %) compared to the provincial average which is normally
in the 5-12 % range. For the horizon year of 1993, traffic composition was taken similar
to the base year plus an additional 20 heavy trucks per day. The directional split for
analysis purposes was taken as 50/50.
4.2
4.2.1
Highway 63 Simulation Model
The Simulation Model
The TRARR (TRAffic on Rural Roads) two-lane highway simulation model developed by
Robinson (1980) at the Australian Road Research Board (ARRB) was used to determine
the impact of heavy trucks on traffic flow. The TRARR model was selected for simulation
purposes because of its flexibility and the fact it is particularly user friendly and well
documented (Hoban, Fawcett, and Robinson, 1985).
The TRARR model can be used to simulate traffic operations on a road in detail
and to investigate the effects of changes in road traffic characteristics. By changing the
traffic characteristics, the user can investigate the effects of increased volumes, more heavy
trucks, or changes in vehicle size and power. Auxiliary lanes may be added on one or both
sides, so that climbing lanes, passing lanes, and three or four-lane roads may be modelled.
TRARR allows passing restrictions, auxiliary lanes, horizontal and vertical curves, and
variable sight distances to be studied. Vehicles may vary in performance, and drivers may
vary in th~ir hehavior when un impeded, following other vehicles, overtaking or merging
The simulation process used in TRARR has the following five main components:
1. read in data on the road, traffic vehicle and driver description, and observing points
along the road;
2. generate a traffic stream with volume, composition, and directional split as specified
in the input files;
3. simulate the movement of each vehicle along the road;
4. record details of traffic characteristics passing certain points or over certain intervals
of roads; and
5. print, plot, or sumnlarize the observed traffic behavior.
Traffic generation with TRARR is based on random (or Monte Carlo) selection of
values from specified distribution of arrival times, speeds, and vehicle types. This is the
only Monte Carlo aspect of the TRARR model since vehicle behavior along the simulated
road is completely determined by vehicle/driver characteristics, and is not subject to any
probabilistic decision-making. Within TRARR, a new vehicle is generated as the preceding
vehicle enters the simulated road segment. Alternatively, it is possible to bypass the traffic
generation routines and read in a file of traffic from outside the program such as a real
traffic data or a separate traffic generation program. The progress of each vehicle along
the road is reviewed at frequent intervals and changes made according to the current state
of the vehicle. Decision rules for overtaking, merging, catching up, and other maneuvers
are determined by the vehicle/ driver characteristics supplied as input. The simulation
cycle also includes a number of book-keeping tasks to maintain and update lists of vehicle
positions and relative order in each lane and between lanes.
4.2.2
Input Data
A major feature of TRARR is its extensive use of input data files, allowing the user to
vary many aspects of the simulation without altering the computer program. The four
main files are as follows:
1. The Road File -
this file gives details of barrier lines, sight distance, curves, and
auxiliary lanes for every 100 m interval in both directions. The road file was created
from data taken from videologs and as-built plans of two sub-sections of Highway
63. Passing sight distance and barrier line information was obtained during field
trips.
2. The Vehicle File - contains approximately 60 vehicle/ driver characteristics for
18 vehicle types. For purposes of this project, a modified Vehicle File was assembled
for 6 vehicle categories to emulate the vehicle/ driver profile characteristics of traffic
on Highway 63.
3. The Traffic File -
traffic composition is described as a rnix of 6 vehicle categories.
Volumes, speeds, and directional splits are specified for each category for existing
and future conditions.
4. The Observation File -
this file allows the user to specify at what points along
the simulated road segment the model should output traffic performance indicators.
The standard observation routine outputs data at a spacing of 1000 m. This spacing
was decreased to 100 m at the start and end of proposed passing lanes in order to
provide detailed information on the diverging and merging processes.
4.2.3
Special Consideration for Loaded and Unloaded Trucks
Unlike traffic composition on rural highways under normal circumstances, the traffic and
vehicle files for Highway 63 had to be further modified in order to reflect the fact that heavy
trucks are loaded in northbound direction and, almost always, unloaded in the southbound
direction. This modification has been achieved by splitting the heavy truck volumes into
north and southbound and assigning 100 % of the northbound trucks to "loaded truck"
category while 0 % of the southbound trucks were assigned to this category. On the other
hand, 100 % of the southbound trucks were assigned to "unloaded truck" category while
o % of the northbound trucks we~e assigned to this category.
4.3
The Output File
The following outputs are given for each direction of travel and each observation point:
overtakings commenced, mean speed and standard deviation, percent vehicles following,
and mean speed by category of vehicle. A summary is also provided for each direction
including, for each vehicle type: mean and standard deviation of travel time, journey
desired and unimpeded speed, percent time spent following, overtakings of and by each
vehicle type, and fuel consumption.
4.4
Simulation Results
Table 5 provides a summary of simulation model runs for two-way volumes ranging from
50 veh/h to 300 veh/h in increments of 50 veh/h. The road files used for the runs
were two sections of the existing highway. The table shows the percentage of time spent
following and the number of overtakings, by direction and total two-way. Percent time
spent following is shown for section 4 in the northbound and southbound directions in
Figures 8 and 9 respectively.
In the northbound direction percent following reaches a maximum of approximately
30 % at the 6 km position as shown in Figure 8. The impact of climbing lanes on percent
following at the 6 km and 13 km position of section 4 is clearly evident. For example, at
a volume of 300 veh/h percent following is reduced approximately 15 percent at the 6 km
position.
Figure 9 shows that percent following reaches a maximum value of approximately 35 %
in the southbound direction. The impact of climbing lanes at the 3 km and 11 km position
is approximately 15 % reduction in percent following.
4.5
Discussion of Simulation Results
Percent time spent following at a volume of 300 veh/h corresponds to level of service B as
determined by the 1985 Highway Capacity Manual (HCM). For purposes of determining
level of service using the 1985 HCM procedures, the terrain was classified as rolling and
22 % no passing zones were used.
Percent time spent following is not the only measure of level of service of a highway.
Other performance indicators such as: number of overtakings, percent time delayed, capacity utilization, and speed should be used in conjunction with percent following. Morrall
and Werner (1988) have suggested a new measure- Overtaking Ratio, defined as the ratio of overtakings on an existing two-lane highway to that on a four-lane highway. For
the highway, terrain, and traffic conditions, the achieved to desired overtaking ratio was
determined to be 0.70. The ratio has been defined by Morrall and Werner (1988) as
follows:
Achieved/Desired Overtaking ratio
where,
=
AO
DO
(11)
AO
Achieved Overtakings: the total number of overtakings for a given
two-lane highway;
DO
Desired Overtakings: the total number of overtakings for a four-lane
highway with vertical and horizontal geometry similar to the given
two-lane highway.
In other words, the overtakings achieved on Highway 63 for the conditions stated are
approximately 70 percent of those possible if Highway 63 were a four-lane highway. If
the volume of traffic were doubled to 600 veh/h, the overtaking ratio would drop by
approximately 50 percent to 0.35. It is noted that the achieved/desired overtaking ratio
decreases at a much faster rate than percent time delay increases through the ranges of
level of service which are most critical for a two-lane highway, namely, the mid-range of
B to the mid-range of C.
In summary, heavy trucks will not have an adverse impact on level of service at a
volume of 300 veh/h. Level of service for this project has been evaluated by three independent measures, namely, simulation, achieved/desired overtaking ratio, and the 1985
Highway Capacity Manual.
5
Conclusions
The application of the Alberta SMV model and the TRARR simulation has assisted in
confirming the initially proposed location of passing/ climbing lanes and truck turnouts as
the most appropriate engineering solution(s) to enhance traffic operation. The models have
also provided the means to evaluate appropraite engineering solutions for other segments
of the HWL corridor.
References
(1] Botma, H. (1988) "Effects on Traffic Operation of a Slow Moving Vehicle on Two-lane
Rural Roads" , a paper presented at the 1988 TRB meeting.
[2] Hoban, C.J., Fawcett, G.J. and Robinson, G.K. (1985) "A Model for Simulating Traffic on Two-Lane Roads: User Guide and Manual for TRARR Version 3.0" , Australian
Road Research Board, Technical Manual ATM No. lOA.
[3] Lay, M.G. (1985) "Source Book for Australian Roads" , Second Edition. Australian
Road Research Board.
[4] Morrall J.F. (1989) "Highway 63 Traffic Operation Analysis During Periods of Expansion of Oil Sands Extraction Plants" , Alberta Transportation and Utilities, Draft
Report.
[5] Morrall, J.F. and Werner, A. (1988) "Measuring Two-Lane Level of Service by Overtakings". Institute of Transportation Engineers Proceedings.
[6] Transportation Research Board (1985) "Highway Capacity Manual- 1985" , Special
Report 209, National Research Council, Washington, D.C.
TABLE 1
Passing for Various Module and Road Widths
Module Width
(ft)
9.1 (30)
7.3 (24)
5.2 (17)
m
RAU 208
No
No
No
Road Standard
RAU 209 RAU 211
No
No
No
Yes
Yes
Yes
RAU 213
Yes
Yes
Yes
TABLE 2
Total Delay (Hours), TD
Truck
Turnout
Spacing (Km)
13.50
14.20
14.60
16.10
16.60
21.70
22.90
ALBERTA Model
Q=100
Q=200
0.74
1.39
0.82
1.52
0.88
1.63
1.08
1.95
2.0i
1.14
3.51
1.94
2.14
3.90
Directional Split
= 50 %
DELFT Model
Q=100 Q=200
0.72
1.48
0.78
1.63
0.83
1.73
2.12
1.02
1.07
2.25
1.80
3.83
2.05
4.26
TABLE 3
Affected Vehicles and Average Delay Per Vehicle (min)t
Truck
Turnout
Spacing (Km)
13.50
14.20
14.60
16.10
16.60
21.70
22.90
t Flow
ALBERTA Model
Aff. Veh Delay/Veh
4.17
20
20
4.56
4.89
20
20
5.85
24
5.18
27
7.80
31
7.55
= 200 Veh/h,
Directional Split
DELFT Model
Aff. Veh Delay/Veh
18
4.93
19
5.15
19
5.46
19
6.69
22
6.14
29
7.92
30
8.52
= 50 %
TABLE 4
Traffic Composition: Highway 63
Vehicle Type
Passenger Cars
Recreational Vehicles
Buses
Heavy Trucks
Single Unit Trucks
Percent
74%
4%
1%
19%
2%
Total
100%
TABLE 5
Simulation Results-Existing Highway
Highway
Section
2-Way
Volume
(Veh/h)
Section 4
(63:04)
50
100
150
200
250
300
4.0
7.3
11.9
14.2
17.4
20.5
7.5
7.9
12.2
17.0
23.2
28.2
5.7
7.6
12.1
15.7
20.1
24.3
5
28
70
126
231
288
5
29
73
140
197
284
10
57
143
266
428
572
Section 5A
(63:05A)
50
100
150
200
250
300
4.5
8.4
12.7
13.2
21.4
24.7
5.9
7.8
10.5
11.7
19.0
25.3
5.2
8.2
11.6
12.4
20.3
25.0
6
23
55
88
165
201
3
26
53
101
132
213
9
49
108
189
297
414
Percent Following
SB 2-WAY
NB
Overtakings
NB SB 2-WAY
\
Oslo Plant Site
;~1~11~!1~1~j~~ McMurray
')
)
,..'
4>
Ii
Fig.1 HighIWide Load Corridors
(Edmonton to Ft McMurray)
Legend:
••••••••
Relative scale
HighIWide Load Corridor
Primary Highway
Secondary Road
Edmonton City Limits
April 18.1989
Ul
~
0
.,-l
Ul
~
Cl)
S
.,-l
0
-
Cl)
,.....;
;::I
'"'d
0
)(
0
;:s
E
N
N
N
-10
CD
0
0
C\I
S
;::I
S
.,-l
~
C'd
:::F.:
N
Cl)
"';::I"'
be
.,-l
~
I
j-
(XDW)
,~
;
I
t
I
I
*- - - - - - -
87.5
100.0
25:1 TAPER
PARALLEL LANE
87.5
PARALLEL LANE
25:1 TAPER
TRUCK TURNOUT "FOR NORTH BOUND LANES
87.5
-25:1 TAPER
.
__ 30.Q_
100.0
- PARALLEL LANE
100.0
.. .30.0...
-
100.0
..
PARALLEL LANE
------------------------------------------------------------------------------
87.5
-
.-
---•
----- --------------_._._._._._.-._._._._.-,_._---_._._-_._._._._._-_._.----_._._._.-._._._._._._._.-._._._._._._._._._._._._._-_.TRUCK TURNOUT FOR SOUTH BOUND LANES
Figure 4.
..
Truck Turnout Dimensions
~-~-----
Figure 5.
Comparison of the Simplified ALBERTA &
DELFT Slow Moving Vehicle Models
Flow
=
200 Veh/h (50/50 Split)
1
o
-
Simplified ALBERTA
-- DELFT
8
~
Ul
L..
:J
0
:::c
~
6
>.
0
-
Q)
Cl
4
J
;
.....-
0
.......
0
I-
2
o
~I----------~--------~----------~--------~--------~
10
13
16
19
Roadside Turnout Spacing (Km)
.,
22
25
1
0
I
---
10
13
16
19
Roadside Turnout Spacing (Km)
~
22
25
Figure 7.
Comparison of the Simplified ALBERTA &
DELFT Slow Moving Vehicle Models
Flow = 200 Veh/h (50/50 Split)
2
o
- Simplified ALBERTA
-- DELFT
~
.£
6
1
5
..c
CD
<:
>.
o
Q)
1
0
--- .
Cl
Q)
0\
e
CD
5
~
o .
I
10
13
16
19
Roadside Turnout Spacing (Km)
..
22
25
....
100
90
80
70
(!)
z
~
9..J
60
~ 50
LEGEND
---------.
Volume =100 veh/h
Volume = 200veh/h
--------
Volume
tZ
lI.J
U
a:::
~
40
=300veh/h
30
~,.\
/'"
20
10
.",.-.,..,r.... ___ - - - -
/
, / ;,--~,
~-,'"
~
o'
o
I
I
I
',-----~IL-
'--
---*
;~I
~
-.1.
I I,
rr <.._ , JII \ _
A ;",,-,\ ~ ,;---_ J I "
/>~~ I ' - - , \' "/'
1\1"
~v
,>~.. ~1' ~\~
\-.J'V
; . . . . .--.
//
I';V",; 1
--'
;
~
~
\;
~-------..;\
;
....
JIlt!
----...
.
~"
~...
,;
i
~
~
=:n ''%"
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 20 21
POSITION (km)
Figure 8.
..
Percent Following vs. Position in the Northbound Direction (Highway Section No. 4)
22 23
100
90
80
70
C!)
z
~ 60
LEGEND
...J
...J
Volume
~ 50
~..",.,,.,~,.,,,,,...,~
--------
IZ
w
U
40
W
-J.. ~
1\
I"
v
~.;
o
'"
"' / -- -...-. . ......,__
r---,,___,-----......_..........,,~~"
I
'
I
' . .____
..."
V',
L---------__
'"
,
r
----_
,-"'_
~,
--~... _
,
. .. ,---"
10
Volume
Volume = 300veh/h
",'"
\
20
=100veh/h
=200veh/h
, ,.""...............,
et:
a. 30
.
'i
\
"
o
2
3
4
5
6
7
8
9
J,,-.
,r
11'
_ _ _ _ _ -.-11'
~
10
-
_ , ;
-
___ _
-,
I1
12
13
14
15
II'~
_
___
16
17
18
11' 11' 11'
19 20
POSITION (km)
Figure 9.
.
Percent Following vs. Position in the Southbound Direction (Highway Section No. 4)
"
' -' -
21
22 23
SESSION 9 - WEIGHTS, DIMENSIONS AND PRODUCTIVITY
Chairman:
Keith Walker, Interjurisdictional Committee on Vehicle Weights
and Dimensions
Speakers
1.
2.
Economics of Vehicle Weights and Dimensions in Canada
L. Sims, N.A. Irwin, IBI Ltd., Toronto
The Benefits of 62.5 tonne, 25 m B Trains in Alberta
T. Fredericks, Economy Carriers, Alberta
3.
The Economic Benefits of Long Combination Vehicle Operations
M. Rice, Transystems, Black Eagle, Montana
4.
The Economics of Road Train Operations in Australia
J. Pamell, Pamell Transport, Australia